Vol. 175, No. 4

Transcriptional Analyses and Mapping of the ospC Genein Lyme Disease Spirochetes

RICHARD T. MARCONI,* D. SCOFT SAMUELS, AND CLAUDE F. GARON

Laboratory of Vectors and Pathogens, Rocky Mountain Laboratories, National InstituteofAllergy and Infectious Diseases, Hamilton, Montana 59840

Received 6 October 1992/Accepted 1 December 1992

In Lyme disease spirochetes, the ospC gene encodes a 22.7-kDa protein referred to as either the pC or theOspC protein. Using a variety of electrophoretic approaches followed by Southern blotting and probing witholigonucleotide probes, we mapped the ospC gene to a circular 26-kb plasmid. The ospC gene represents thefirst gene to be mapped to a circular plasmid in Lyme disease spirochetes. The occurrence of this gene in isolatesbelonging to each of the three Lyme disease-associated species, Borrelia burgdorfieri, Borrelia garinii, and theVS461 group, was evaluated. The ospC gene was found to occur in all 21 isolates tested from each of the threespecies. Differential hybridization with a series of ospC probes in both Northern (RNA) and Southern blotanalyses demonstrated that there is sequence variability in the ospC gene among isolates. While the gene wasfound to be present in all isolates, not all actively transcribed the gene. Transcriptional start site analysessuggest that the gene may be under the control of multiple promoters that are highly similar in nucleotidesequence.

The organization of the genetic material among membersof the genus Borrelia is unusual among eubacteria in that thegenetic material is distributed between a linear chromosomeand a series of both linear and circular plasmids (4, 5, 10).Although variable, Borrelia plasmids account for a signifi-cant fraction of the total genetic material. In Lyme diseaseborreliae, all genes mapped to date have been found to resideon either the 1,000-kb linear chromosome (12, 28) or, in thecase of the ospAB operon, on a linear plasmid (5, 6, 8) thatranges in size from approximately 49 to 59 kb (4, 5, 13, 23).In general, the coding capacity of the plasmids remainslargely undefined. It seems likely that in addition to ospAB,other genes will also be mapped to plasmids.

Recently, a third outer surface protein (OspC) in Lymedisease spirochetes was identified, and its gene was se-

quenced (11). The physiological role or pathological signifi-cance of OspC has not yet been defined. Although somewhatvariable among isolates, the molecular mass of OspC isapproximately 22.7 kDa. OspC appears distinct from the22-kDa proteins described by Luft et al. (14) and Simpson etal. (26). The level of expression of OspC has also been foundto vary among isolates. Wilske et al. reported that OspC isexpressed as a major protein in approximately 45% ofEuropean isolates and only rarely in North American iso-lates (31). For the three spirochete species, Borrelia burg-dorfeni, Borrelia garinii, and group VS461, known to beassociated with Lyme disease (2, 15-17, 19, 29), the pres-ence, degree of conservation, and map location of the ospCgene have not yet been addressed.

In this study, we have mapped the location of the ospCgene, determined its transcriptional start site(s), and com-

pared the expression of the gene among isolates of Lymedisease spirochetes. The results presented here demonstratethat the ospC gene is carried on a 26-kb circular plasmid. TheospC gene represents the first Borrelia gene to be mapped toa circular plasmid. Transcriptional start site analyses haverevealed that the control region of the gene is similar among

* Corresponding author.

isolates of Lyme disease spirochetes and that the gene maybe under the control of multiple promoters. Northern (RNA)blot analysis has also demonstrated that the transcriptionalexpression of the ospC gene varies among isolates.

MATERMILS AND METHODS

Cultivation of Borrelia isolates. The isolates investigated inthis study are described in Table 1. All isolates were previ-ously identified to the species level by 16S rRNA sequenceanalysis (1, 16, 17), through the use of 16S rRNA-directedspecies-specific oligonucleotide probes (19) or polymerasechain reaction primer sets (15), by DNA-DNA hybridization(2, 20), or by multilocus enzyme electrophoresis (9). Thespecies nomenclature used in this paper is that proposed byBaranton et al. (2). Bacterial isolates were cultivated inBSKII medium at 34°C (3), harvested by centrifugation, andwashed twice with phosphate-buffered saline prior to use insubsequent experiments.

Southern blot analysis of the ospC gene. DNA was isolatedas previously described (7). Whole-cell DNA was fraction-ated by electrophoresis in 0.35% agarose gels with TAEbuffer (40 mM Tris-acetate, 1 mM EDTA [pH 8.5]) at 0.4V/cm for 48 h. Fractionated DNA was transferred to Gene-Screen membranes by use of the VacuGene vacuum blotsystem as instructed by the manufacturer (Pharmacia-LKB).The transferred DNA was fixed to the membranes by UVcross-linking with a Stratalinker (Stratagene) as instructedby the manufacturer. ospC hybridization probes were syn-thesized on the basis of the ospC sequence previouslypresented for the Lyme disease isolate PKo (11). Both Adamet al. (1) and Wallich et al. (28) have identified this isolate as

belonging to the VS461 group of Lyme disease spirochetes.Oligonucleotide probe sequences are presented in Table 2.All sequences given in this paper are listed 5' -* 3'. Oligo-nucleotide probes were 5' end labeled with [_y-32P]ATP byuse of polynucleotide kinase by standard methods. Radiola-beled probes were purified from unincorporated label bypassage through G25 spin columns as instructed by themanufacturer (Boehringer Mannheim). The blots were pre-

926

JOURNAL OF BACTERIOLOGY, Feb. 1993, p. 926-9320021-9193/93/040926-07$02.00/0Copyright X) 1993, American Society for Microbiology

on March 11, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ospC GENE MAPPING AND TRANSCRIPTIONAL ANALYSES 927

TABLE 1. Borrelia isolates used in this study

Isolates Geographic Biological sourcea

Borrelia burgdorferi25015 New York Ixodes damminiIllinois 1 Illinois Mouse1352 Texas Amblyomma americanum20004 France Icodes ricinusIP2A France Human CSFCA12 California Icodes pacificus3028 Texas Human pus

Borrelia ganrniiG25, VS102 Sweden Ixodes ricinusG1 Germany Human CSFR-IP90 Russia Ixodes persulcatus20047 France Ixodes ricinusN34 Germany Ixodes ricinus

VS461 groupU01, ECM1, UM01 Sweden EM (human)R-IP21, R-IP3 Russia Ixodes persulcatusVS461 Sweden Ixodes ricinusB023 Germany Human skinJi Japan Ixodes persulcatusa CSF, cerebrospinal fluid; EM, erythema migrans.

hybridized in hybridization buffer (16) for 1 to 4 h at 370C.After prehybridization, the buffer was removed and radio-active probes at 10 ng/ml in fresh hybridization buffer wereadded to the hybridization bottles. Hybridization was con-ducted for 16 to 24 h at 370C. The blots were washed twicein 2x SSC (lx SSC is 0.15 M NaCl plus 0.015 M sodiumcitrate)-O.1% sodium dodecyl sulfate (SDS) at 370C (10 mineach) and once for 1 h in O.lx SSC-0.1% SDS at 370C,wrapped in cellophane, and exposed to Kodak XAR5 film at-700C with intensifying screens.Two-dimensional agarose gel electrophoresis. For determi-

nation of the conformation of the DNA molecule carryingthe ospC gene, two-dimensional gel electrophoresis wasperformed (21, 22). DNA from isolates R-IP21 and VS461was electrophoresed in 0.35% agarose gels in the firstdimension for 18 h at 0.4 V/cm with TAE buffer. The gelswere then equilibrated in second-dimension running buffer(15 pM chloroquine in TAE buffer) by soaking for S h. Afterthe gels were rotated 900 relative to the direction of electro-phoresis in the first dimension, electrophoresis was per-formed for 18 h at 0.4 V/cm. Prior to ethidium bromidestaining, the gels were soaked in H20 for several hours toremove the chloroquine. Fractionated DNA was transferred

TABLE 2. Oligonucleotide hybridization probe orprimer sequences

Probe or Sequence (5' 3')' Targetprimer regionb

pC13 GTC-ATT-AAT-ATC-GCA-CTT-AAT-G 13 to 35pC52 CCC-TGA-ATT-ATT-ACA-AGA 52 to 66pC67 TGC-AGA-ATC-CCC-ACC-TTT-CCC 67 to 87pC609 TGG-ACT-TTC-TGC-TAC-AAC-AGG 609 to 629nc2O ATG-AAT-TAT-AGT-CCA-ACA-AT -40 to -20

a All probes or primers complement the coding (plus) strand.b The numbers given are relative to the translational start codon of the ospC

sequence previously presented for VS461 group isolate PKo (11).

to GeneScreen membranes for hybridization analysis withospC oligonucleotide probes as described above.

Transcriptional start site and Northern blot analyses. Totalcellular RNA was isolated as previously described (25). Allreagents were either prepared with diethylpyrocarbonate-treated H20 or treated directly with diethylpyrocarbonate bystandard methods. The transcriptional start site(s) for theospC gene was determined by reverse transcriptase primerextension. Fifteen micrograms of total cellular RNA wasincubated with 2 pmol of 5'-end-labeled primer pC13, pC52,pC67, or nc2O in 10 RI1 of annealing buffer (100 mM KCI, SmM Tris-HCI [pH 7.5]) at 850C for 4 min and then at 30'C for4 to 6 h. Fifteen microliters of reverse transcription mixture(0.9 mM each dATP, dTTP, dCTP, and dGTP; 83 mMTris-HCI [pH 8.3]; 8 mM dithiothreitol; 8 mM MgCl2; 83 pugof bovine serum albumin per ml) and 50 U of avian myelo-blastosis virus reverse transcriptase were added, and thereaction mixtures were incubated for 1.5 h at 47°C. Fordigestion of the RNA, 1 ,u1 of 0.5 M EDTA and 1 ,u1 of RNase(0.5 Rg/pRl) were added and the mixture was incubated for 0.5h at 37°C. One volume of 2.5 M ammonium acetate wasadded to each tube, and each was extracted with an equalvolume of phenol-chloroform-isoamyl alcohol. The exten-sion products were precipitated with ethanol, washed with70% ethanol, vacuum dried, resuspended in 10 ,u1 of forma-mide loading buffer (50% formamide, 10 mM EDTA, 0.025%bromophenol blue, 0.025% xylene cyanol), heated at 85°Cfor S min, and loaded onto 7% polyacrylamide sequencinggels (8 M urea, 83 mM Tris-borate, 1 mM EDTA [pH 8.0]).The approximate sizes of the extension products were de-termined by use of a DNA sequencing ladder. After electro-phoresis, the gels were transferred directly to 3MM paper(Whatman) and exposed to Kodak XAR5 film for 3 h.Northern blot analysis was used to assess the expression

of the ospC gene and to determine the nucleotide length ofthe transcript. Fifteen micrograms of RNA sample waselectrophoresed through 1.2% agarose-formaldehyde gels aspreviously described (27). The RNA was transferred toGeneScreen membranes and hybridized as described abovefor the Southern blot analysis. The length of the ospCtranscript was determined by use of an NA2 nucleic acidanalyzer as instructed by the manufacturer (Bethesda Re-search Laboratories).

RESULTS AND DISCUSSION

Mapping of the ospC gene to a circular plasmid. In thisstudy, we sought to map the ospC locus in a variety ofisolates of each of the three Lyme disease-associated spiro-chete species (2, 9, 15-17, 19, 20, 29). Total DNA from theisolates listed in Table 1 was isolated and fractionated in0.35% agarose gels. After Southern transfer, the blots wereprobed with an ospC oligonucleotide probe (pC13). Two orthree hybridization bands that differed in intensity andrelative ratio among isolates were observed. The ethidiumbromide-stained gel and the corresponding Southern blot ofrepresentative isolates are presented in Fig. 1. The detectionof more than one band suggested that the probe was eitherhybridizing with an ospC gene carried on multiple DNAspecies or cross-hybridizing with non-ospC but closely re-lated sequences. To differentiate between these possibilities,we used a second probe (pC609), which complements aregion in the 3' end of the ospC gene, to rescreen some of theblots. This probe hybridized with the same bands as thepC13 probe. A third probe (pC67) also yielded identicalhybridization results. Both the pC67 and the pC609 probes,

VOL. 175, 1993

on March 11, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

928 MARCONI ET AL.

A B1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9

48.5 -

33.5 -II et*t-

24.5--III-

17-7;t X!

FIG. 1. Agarose gel fractionation of Lyme disease spirochete DNA and Southernblot analysis of the ospC gene.TotalIDNA isolated fromthe three Lyme disease-associated organisms was fractionated in 0.35% agarose gels, stained with ethidium bromide (A), transferred toGeneScreen membranes, and probed with the pCl3 probe (B) as described in the text. DNA from the following respective isolates wasanalyzed: lanes 1 and 2, B. garinii VS102 and R-1P90; lanes 3 and 4, B. burgdorferi 20004 and IP2A; and lanes 5 to 9, VS461 group isolatesR-IP21, VS461, R-IP3, ECM1, and UMOL. DNA molecular size markers are given in kilobases. The three plasmid forms are indicated by I(supercoiled), II (relaxed), and III (linearized).

however, did not hybridize with the ospC gene in all isolates,suggesting sequence variability in this gene (see below).Nonetheless, the use of multiple probes, all of which hybrid-ized with the same DNA bands, demonstrates that themultiple hybridization signals observed were due to thedetection of multiple ospC-carrying DNA molecules.The detection of multiple hybridizing bands and their

relative migration rates in agarose gels suggested that thegene may be carried on a circular plasmid. Circular plasmidscan be found in three distinct conformational states. Form Iis the supercoiled form. Form II consists of relaxed mole-cules that comigrate with supercoiled plasmids that havebeen nicked. Form III represents plasmids that have beenlinearized by a double-strand break. The hybridization pat-tern suggested a circular plasmid of 26 kb, with the super-coiled form (I) migrating with an apparent size of 12 kb andthe nicked or relaxed form (II) migrating with an apparentsize of 44 kb (Fig. 1B). Furthermore, upon restriction of theDNA from several isolates with EcoRV, Southern blotting,and probing with the pC67 probe, only one hybridizationband was observed (data not shown). These results furthersupport the idea that one copy of the ospC gene was presentand being carried on different conformational forms of thesame plasmid.While the detection of multiple hybridizing bands in

undigested DNA preparations is highly suggestive of acircular plasmid, we further addressed this question by usingtwo-dimensional gel electrophoresis. Linear and circularplasmids, which cannot be distinguished from each other byconventional agarose gel electrophoresis, can be conclu-sively differentiated via this approach. For these analyses,new DNA preparations were gently isolated (i.e., vortexingand pipetting were minimized during isolation) from VS461group isolates R-IP21 and VS461 to reduce shear damage tothe circular plasmids in the DNA preparations. After stan-

dard electrophoresis of the DNA in the first dimension, thegels were equilibrated in chloroquine. Chloroquine interca-lates into the DNA, causing an unwinding of the doublehelix. The electrophoretic mobility of linear molecules is notsignificantly affected by chloroquine treatment. These mol-ecules migrate along an apparent diagonal axis in two-dimensional chloroquine gels. However, the electrophoreticmobility of negatively supercoiled plasmids is retarded bychloroquine treatment and, as a result, these plasmids mi-grate off the diagonal axis during electrophoresis in thesecond dimension (22). Three DNA species were found inethidium bromide-stained gels to migrate off the diagonalaxis, indicating that these molecules were covalently closedand negatively supercoiled (Fig. 2A). The gels were thenblotted and subjected to Southern blot analysis with thepC67 hybridization probe. A hybridization signal was de-tected associated with one of the molecules running off thediagonal axis (Fig. 2B), demonstrating that the ospC gene iscarried on a single negatively supercoiled plasmid. In con-trast to the results of the Southern blot analysis describedabove, only one band was found to hybridize with the probe.Hence, the elimination of vortexing from the DNA isolationprocedure significantly reduced shearing of the circularplasmids and, as a result, only the supercoiled form of theospC-carrying plasmid was observed. The size of the circu-lar plasmid was obtained by measuring the linearized form(Fig. 1) and found to be 26 kb.Occurrence and conservation of the ospC gene in Lyme

disease spirochetes. Wilske et al. have reported that OspC isexpressed in 45% of European isolates and only rarely inNorth American isolates (31). To evaluate the occurrence ofthe ospC gene, we performed Southern blotting of undi-gested whole-cell DNA from a variety of isolates of the threespirochete species associated with Lyme disease. The re-sults obtained with representative isolates are presented in

J. BACTERIOL.

on March 11, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ospC GENE MAPPING AND TRANSCRIPTIONAL ANALYSES 929

A B0

485 -

29.24.0 -17.0

10.1 -

=a

NoaI0(A

2nd dimensionFIG. 2. Two-dimensional gel electrophoresis of DNA from Lyme disease isolate VS461 and Southern blot analysis with the pC67 probe.

Electrophoresis of DNA from isolate VS461 was conducted in the first dimension in 0.35% agarose gels. The gels were then equilibrated in15 puM chloroquine, rotated 90°, and electrophoresed in the second dimension as described in the text. The DNA was visualized by ethidiumbromide staining (A) and then transferred to GeneScreen membranes for Southern blotting as described in the text with the radiolabeled pC67oligonucleotide probe (B). In both panels, the position of the loading well is indicated by a circle, and that of the 26-kb plasmid is indicatedby an arrow. The direction of migration in each dimension is also given. DNA molecular size markers are given in kilobases.

Fig. 1. Detection of the ospC gene in a given isolate wasfound to be dependent on the probe used. While the pC609probe was found to exhibit differential hybridization withdifferent isolates, this probe was not used to screen allisolates and hence will not be discussed further. Most B.burgdorferi isolates and B. garinii G1 and R-IP90 werehybridization negative with the pC67 probe but hybridizationpositive with the pC13 probe. B. burgdorfen 25015 andIllinois 1 were exceptions in that they hybridized with boththe pC67 and the pC13 probes. We have demonstrated theperipheral relationship of these isolates to others of B.burgdorferi via 16S rRNA sequence analysis (15). In con-trast to B. gannii R-IP90 and G1, VS461 group isolates andB. gannii G25, 20047, and VS102 were hybridization positivewith both the pC13 and the pC67 probes. Hence, while thephenotypic expression of the ospC gene appears most pro-nounced in European isolates, the ospC gene itself is presentin all three Lyme disease-associated spirochete species.Furthermore, it consistently maps to an approximately 26-kbcircular plasmid in all isolates tested. The variability inhybridization among isolates indicates that the ospC genesequence may be somewhat variable. A similar variabilityhas been noted for outer surface protein genes ospA andospB (7, 13, 18, 30).

Northern blot analysis of the ospC transcript. To determinein which isolates the ospC gene was being transcriptionallyexpressed, we performed Northern blot analysis with thepC13 and pC67 hybridization probes. Consistent with theSouthern blot analysis discussed above, the detection of theospC transcript in a given isolate was dependent on theprobe used. ospC transcript size was constant among iso-lates, being approximately 700 to 730 nucleotides (Fig. 3).The ospC coding sequence alone (excluding 5' and 3' non-coding sequences) would generate a 636-base transcript.While the ospC gene was detected by Southern blot analysisin all isolates with either the pC13 or the pC67 probe, we didnot detect the transcript in several isolates by Northern blotanalysis with these same probes. B. burgdorferi CA12 and

1352, B. garinii N34, and VS461 group isolates J1, R-IP21,and UM01 were all found to carry the ospC gene but not toexpress it under the growth conditions used in these exper-iments. The rRNA bands in the hybridization-negative prep-arations were found to show good integrity via ethidiumbromide staining, providing indirect evidence that the lack ofhybridization was not the result of template degradation.Furthermore, using these same RNA preparations, we wereable to detect the ospAB transcript. Factors associated withthe transcriptional activation of the ospC gene have not yetbeen defined. In addition, several VS461 group isolates thatproduced the ospC transcript did not exhibit detectableamounts of OspC protein, as determined by inspection ofCoomassie-stained sodium dodecyl sulfate-polyacrylamidegel electrophoresis whole-cell protein profiles (data not

h CO b c- ffi o C

cm m m m2Mmmm 5 m . m!P-~~ ~~~~~~V 0. A::. C4.X

23S. .Lyt.E

FIG. 3. Detection of the ospC transcript by Northern blot anal-ysis. Fifteen micrograms of total cellular RNA was fractionated in1.2% agarose-formaldehyde gels, vacuum blotted onto GeneScreenmembranes, and hybridized with the pCl3 probe as described in thetext. The species designation for each isolate is indicated by theprefixes Bb for B. burgdorferi, Bg for B. garinii, and (VS) for VS461group isolates. The positions of 23S and 16S rRNAs (approximately2,900 and 1,540 bases, respectively) are indicated.

VOL. 175, 1993

"W"0-

.l

on March 11, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

930 MARCONI ET AL.

9# >"

m m >

53-50-

45 -

40 -

35 -

30 -

25 -

20 -

FIG. 4. Transcriptional start site analyses by reverse tran-scriptase primer extension. All methods were as described in thetext. The pC52 primer was used in the primer extension reactions.The lanes (left to right) contain the extension products obtainedwhen total RNAs (15 pLg) from B. burgdorferi 3028, B. garniiVS102, and VS461 group isolate ECM1, respectively, were used as

templates. The numbers on the left indicate the approximate numberof bases 5' of the translational start codon for each major primer

extension termination site, as determined by electrophoresis of asequencing ladder in adjacent lanes.

shown). OspC expression in these isolates may be subject tosome form of translational control. Alternatively, there maybe mutations in the gene that result in premature transla-tional termination. It has been demonstrated that the pheno-typic expression of the ospA and ospB genes in some isolatescan be influenced by the number of in vitro passages (24).While in this study we did not systematically monitor OspCexpression over a large number of passages in a givenisolate, isolates of both high and low passage numbers wereused. No correlation between passage number and OspCexpression was observed.

Transcriptional start site analyses of the ospC gene. Toidentify the putative promoter(s) of the ospC gene, weperformed transcriptional start site analyses by using reversetranscriptase primer extension. By using two independentprimers (pC13 and pC52), the same primer extension termi-nation sites were observed in all isolates which had been byNorthern blot analysis to be expressing the ospC gene. Themajor termination sites mapped approximately 18 to 22nucleotides upstream from the translational start codon. Theresults obtained with a representative isolate of each of thethree Lyme disease-associated species and the pC52 primerare presented in Fig. 4. An appropriately spaced consensuspromoter sequence (5'-TTG-AAA-A) that we refer to as P1and a consensus TATA box sequence (5'-TATTAA) wereidentified upstream (on the basis of the ospC sequence forisolate PKo) (Fig. 5). Hence, the transcripts derived from theuse of promoter P1 possess a 5'-untranslated leader se-quence of approximately 20 nucleotides, with the T residuelocated 20 bases from the translational start codon serving asthe major transcriptional start site (TS1). Using the RNAsecondary structure analysis program of Zucker (containedin the PCGENE software package) (32), we analyzed se-

-110 P3 P2 *aat-TTG-AAA-Aat-tat-tta-aaa-tat-tac-tcT-TGA-AAA-att-

-71 *Piatt-ttt-ttt-ata-taa-aaa-aTT-GAA-AAg-aaa-aat-tgt-tgg-

-32 * +1act-att-aat-tca-taa-aaa-gga-ggc-aca-aat-taATG

FIG. 5. Putative ospC control regions. -35 promoter regions(P1, P2, and P3) inferred from reverse transcriptase primer exten-sion experiments are indicated by capital letters, and -10 TATAbox sequences are underlined. Nucleotide position numbering iswith respect to the first base (+1) of the translational start codonindicated by italics. Transcriptional start sites are indicated by anasterisk over the appropriate nucleotide. The sequence presented isfrom VS461 group isolate PKo (11).

quences 5' of the translational start codon for regions ofsecondary structure that could conceivably be responsiblefor the termination sites observed in the primer extensionanalysis. However, significant secondary structure elementswere not observed, supporting the identification of TS1 asthe major transcriptional start site.Through a consensus promoter sequence homology

search, Fuchs et al. (11) identified a putative promoterupstream of P1 that we refer to as P2. This sequence is notoptimally spaced to serve as the promoter for TS1. How-ever, when both the pC13 and the pC52 primers were used,a second transcriptional start site (TS2) was observed up-stream of the TS1 site, implicating active transcription froma promoter 5' of P1 (Fig. 4). In B. burgdorfen 3028, a minorextension product intermediate in size in comparison withthose derived from P1 and P2 was observed. The sequenceof this product is not appropriately spaced from a consensuspromoter sequence, so this product most likely represents apremature termination product. P2 appears appropriatelyspaced from TS2 to serve as a promoter for this transcrip-tional start site. Visualization of the band corresponding toTS2 required a longer exposure of the separating gel to film(approximately 6 to 10 h) than that required for the detectionof TS1. The exposure presented in Fig. 4 was chosen since itallowed visualization of both transcriptional start sites with-out overexposure of TS1. Both P1 and P2 are identical insequence over a 7-base stretch (5'-TTG-AAA-A) and haveappropriately spaced -10 sequences (Fig. 5). The coresequences of P1 and P2 and their respective -10 sequencesare separated by 15 and 13 bases, respectively. Whiletranscriptional start sites were observed appropriatelyspaced from both P1 and P2, it is possible that only the moredistal promoter, P2, was used and that the shorter transcriptwas the result of an RNA processing event. While the datapresented here are suggestive of active transcription frommore than one promoter, RNA processing cannot be ruledout.Upon analysis of the PKo ospC sequence (11), we noted a

third potential promoter sequence (P3) upstream of andidentical to the 7-base core sequences of P1 and P2. Over a16-base stretch, P3 exhibits 88% sequence similarity to P2.To determine whether this putative promoter sequence wastranscriptionally active, we used a third primer (nc2O) in theprimer extension analysis. This primer complements a se-quence in the ospC gene 5' of TS1. While the 5' end of this20-base primer complements the first 6 nucleotides of tran-scripts derived from the expression of promoter P2, thesetranscripts should not serve as a template for primer exten-sion, since the 3' end of this primer does not hybridize.

J. BACTERIOL.

on March 11, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ospC GENE MAPPING AND TRANSCRIPTIONAL ANALYSES 931

Hence, any extension products should be derived fromhybridization with a transcript that uses a more distal 5'promoter. In two isolates (B. burgdorfen 3028 and B. ganniiVS102), a weak transcription start site, not previously ob-served when primers pC13 and pC52 were used, was de-tected (data not shown). Hence, promoter P3 may be used ata low level in the transcription of the ospC gene in someisolates. The low level of transcription from this promotermay be a reflection of the spacing between the -35 and -10sequences (11 bases). None of the three transcriptional startsites described above was observed in isolate R-IP21, whichwas shown not to produce the ospC transcript by Northernblot analysis, demonstrating that the start sites detectedwere not the result of spurious or nonspecific primer exten-sion from non-ospC transcripts. The role of each promoter,the extent to which each is used, and the factors influencingpromoter selection in vivo are currently undefined.The ospC gene represents the first Borrelia gene to be

mapped on a circular plasmid. The plasmid is approximately26 kb in size. On the basis of hybridization results with avariety of oligonucleotide probes, this gene appears toexhibit some variability in its coding sequence among theLyme disease isolates tested. The organization of the up-stream regulatory regions, however, appears similar amongisolates with a potential multiple-promoter arrangement. Theresults presented here may provide further insight into thegenetic coding capacity of the plasmids found in Lymedisease isolates. Since the plasmids represent approximate-ly 150 kb of genetic material, it is likely that additionalgenes will be localized on plasmids and shown to encodeproducts that may be important in the pathobiology of Lymedisease.

ACKNOWLEDGMENTS

We thank J. Anderson, G. Baranton, A. Barbour, S. Barthold, J.Nelson, 0. Peter, D. Postic, J. Rawlings, P. Rosa, and T. Schwanfor providing isolates; J. Heinemann and J. Hinnebusch for helpfuldiscussions and critical review of the manuscript; and R. Evans andG. Hettrick for providing the graphic work.

REFERENCES

1. Adam, T., G. S. Gassmann, C. Rasiah, and U. B. Gobel. 1991.Phenotypic and genotypic analysis of Borrelia burgdorferi iso-lates from various sources. Infect. Immun. 59:2579-2585.

2. Baranton, G., D. Postic, I. Saint Girons, P. Boerlin, J.-C.Piffaretti, M. Assous, and P. A. D. Grimont. 1992. Delineation ofBorrelia burgdorfen sensu stricto, Borrelia gannii sp. nov., andgroup VS461 associated with Lyme borreliosis. Int. J. Syst.Bacteriol. 42:378-383.

3. Barbour, A. G. 1984. Isolation and cultivation of Lyme diseasespirochetes. Yale J. Biol. Med. 57:521-525.

4. Barbour, A. G. 1988. Plasmid analysis of Borrelia burgdorferi,the Lyme disease agent. J. Clin. Microbiol. 26:475-478.

5. Barbour, A. G., and C. F. Garon. 1987. Linear plasmids of thebacterium Borrelia burgdorferi have covalently closed ends.Science 237:409-411.

6. Barbour, A. G., and C. F. Garon. 1988. The genes encodingmajor surface proteins of Borrelia burgdorfen are located on aplasmid. Ann. N.Y. Acad. Sci. 539:144-153.

7. Barbour, A. G., R. A. Heiland, and T. R. Howe. 1985. Hetero-geneity of major proteins in Lyme disease borreliae: a molecularanalysis of North American and European isolates. J. Infect.Dis. 152:478-484.

8. Bergstrom, S., V. G. Bundoc, and A. G. Barbour. 1989. Molec-ular analysis of the linear plasmid encoded major surfaceproteins, OspA and OspB, of the Lyme disease spirochaete

Borrelia burgdorferi. Mol. Microbiol. 3:479-486.9. Boerlin, P., 0. Peter, A.-G. Bretz, D. Postic, G. Baranton, and

J.-C. Piffaretti. 1992. Population genetic analysis of Borreliaburgdorfen isolates by multilocus enzyme electrophoresis. In-fect. Immun. 60:1677-1683.

10. Ferdows, M. S., and A. G. Barbour. 1989. Megabase-sized linearDNA in the bacterium Borrelia burgdorferi, the Lyme diseaseagent. Proc. Natl. Acad. Sci. USA 86:5969-5973.

11. Fuchs, R., S. Jauris, F. Lottspeich, V. Preac-Mursic, B. Wilske,and E. Soutschek. 1992. Molecular analysis and expression of aBorrelia burgdorfen gene encoding a 22 kDa protein (pC) inEscherichia coli. Mol. Microbiol. 6:503-509.

12. Gassman, G. S., M. Kramer, U. B. Gobel, and R. Wallich. 1989.Nucleotide sequence of the gene encoding the Borrelia burgdor-fen flagellin. Nucleic Acids Res. 17:3590.

13. Jonsson, M., L. Noppa, A. G. Barbour, and S. Bergstrom. 1992.Heterogeneity of outer membrane proteins in Borrelia burgdor-feri: comparison of osp operons of three isolates of differentgeographic origins. Infect. Immun. 60:1845-1853.

14. Luft, B. J., W. Jiang, P. Munoz, R. J. Dattwyler, and P. D.Gorevic. 1989. Biochemical and immunological characterizationof the surface proteins of Borrelia burgdorfeni. Infect. Immun.57:3637-3645.

15. Marconi, R. T., and C. F. Garon. 1992. Development ofpolymerase chain reaction primer sets for diagnosis of Lymedisease and for species-specific identification of Lyme diseaseisolates by 16S rRNA signature nucleotide analysis. J. Clin.Microbiol. 30:2830-2834.

16. Marconi, R. T., and C. F. Garon. 1992. Phylogenetic analysis ofthe genus Borrelia: a comparison of North American andEuropean isolates of Borrelia burgdorferi. J. Bacteriol. 174:241-244.

17. Marconi, R. T., and C. F. Garon. 1992. Identification of a thirdgenomic group of Borrelia burgdorferi through signature nucle-otide analysis and 16S rRNA sequence determination. J. Gen.Microbiol. 138:533-536.

18. Marconi, R. T., M. E. Konkel, and C. F. Garon. Submitted forpublication.

19. Marconi, R. T., L. Lubke, W. Hauglum, and C. F. Garon.1992. Species-specific identification of and distinction betweenBorrelia burgdorferi genomic groups by using 16S rRNA-directed oligonucleotide probes. J. Clin. Microbiol. 30:628-632.

20. Postic, D., C. Edlinger, C. Richaud, F. Grimont, Y. Dufresne, P.Perolat, G. Baranton, and P. A. D. Grimont. 1990. Two genomicspecies in Borrelia burgdorferi. Res. Microbiol. 141:465-475.

21. Pruss, G. J. 1985. DNA topoisomerase I mutants: increasedheterogeneity in linking number and other replicon-dependentchanges in DNA supercoiling. J. Mol. Biol. 185:51-63.

22. Samuels, D. S., and C. F. Garon. 1993. Coumermycin A1 inhibitsgrowth and induces relaxation of supercoiled plasmids in Bor-relia burgdorfeni, the Lyme disease agent. Antimicrob. AgentsChemother. 37:46-50.

23. Samuels, D. S., R. T. Marconi, and C. F. Garon. Submitted forpublication.

24. Schwan, T. G., W. Burgdorfer, and C. F. Garon. 1988. Changesin infectivity and plasmid profile of the Lyme disease spiro-chete, Borrelia burgdorferi, as a result of in vitro cultivation.Infect. Immun. 56:1831-1836.

25. Selden, R. F. 1991. Preparation and analysis of RNA, p.4.4.4.-4.4.6. In F. M. Ausubel, R. Brent, R. E. Kingston, D. D.Moore, J. G. Seidman, J. A. Smith, and K. Struhl (ed.), Currentprotocols in molecular biology. Greene Publishing Associatesand Wiley-Interscience, New York.

26. Simpson, W. J., W. Burgdorfer, M. E. Schrumpf, R. H.Karstens, and T. G. Schwan. 1991. Molecular and immunologi-cal analysis of a polymorphic periplasmic protein of Borreliaburgdorferi. J. Clin. Microbiol. 29:1940-1948.

27. Thomas, P. S. 1980. Hybridization of denatured RNA and smallDNA fragments transferred to nitrocellulose. Proc. Natl. Acad.Sci. USA 77:5201-5205.

28. Wallich, R., C. Helmes, U. E. Schaible, Y. Lobet, S. E. Moter,M. D. Kramer, and M. M. Simon. 1992. Evaluation of genetic

VOL. 175, 1993

on March 11, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

932 MARCONI ET AL.

divergence among Borrelia burgdorferi isolates using OspA, fla,HSP60, and HSP70 gene probes. Infect. Immun. 60:4856-4866.

29. Welsh, J., C. Pretzman, D. Postic, I. Saint Girons, G. Baranton,and M. McClelland. 1992. Genomic fingerprinting by arbitrarilyprimed polymerase chain reaction resolves Borrelia burgdorfeninto three distinct phyletic groups. Int. J. Syst. Bacteriol.42:370-377.

30. Wilske, B., V. Preac-Mursic, G. Schierz, and K. V. Busch. 1986.

Immunochemical and immunological analysis of European Bor-relia burgdorferi strains. Zentralbl. Bakteriol. Mikrobiol. Hyg.Ser. A 263:92-102.

31. Wilske, B., V. Preac-Mursic, G. Schierz, R. Kuhbeck, A. G.Barbour, and M. Kramer. 1988. Antigenic variability of Borreliaburgdorferi. Ann. N.Y. Acad. Sci. 539:126-143.

32. Zucker, M., and P. Stiegler. 1981. Description of the methodused in RNAFOLD. Nucleic Acids Res. 9:133-148.

J. BACTERIOL.

on March 11, 2020 by guest

http://jb.asm.org/

Dow

nloaded from